1. Field of the Invention
The invention relates to an atomic mask useful for a quantum well semiconductor laser utilizing quantum effects which take place in Bohr's atomic radius or smaller, the formation of a pattern in a ultra-micro device such as a single electron transistor, and the formation of a pattern in a recording medium such as CD-ROM. The invention also relates to a method of patterning a substrate with an atomic mask.
2. Description of the Related Art
As semiconductor devices have been fabricated in smaller and smaller sizes, a metal pattern is designed to have a narrower width, and is sometimes required to have an atomic-sized width on the sub-nanometer or nanometer order. Thus, there is a need for forming a fine pattern having an atomic-sized width.
To this end, there have been suggested many methods employing a scanning tunneling microscope (hereinafter, referred to simply as a "STM") . One of such methods has been suggested in I-W. Lyo et al., Science 253, pp. 173, 1991. In the suggested method, a probe of a STM is placed close to a substrate at such a distance as a tunnel current could run therebetween with a voltage being applied across the probe and the substrate. Atoms are desorbed out of the substrate because of the field evaporation effect which takes place when a tunneling current is generated.
In another method suggested by H. J. Mamin et al., "Atomic Emission from a Gold Scanning-Tunneling-Microscope Tip", Physical Review Letters, Vol. 65, No. 19, 1990, pp. 2418-2421, there is used a probe of a STM which is coated with gold atoms by evaporation. The gold atoms are transferred from the probe to a substrate and deposit on the substrate because of the field evaporation effect.
In still another method suggested by M. Baba et al., "Nanostructure Fabrication by Scanning Tunneling Microscope", Japanese Journal of Applied Physics, Vol. 29, No. 12, 1990, pp. 2854-2857, chlorine or fluorine gas is applied onto a substrate, and etching is carried out just beneath a probe of a STM to thereby pattern the substrate. For instance, a metal gas such as WF.sub.6 gas is flowed onto a substrate, and is decomposed just beneath a probe of a STM and deposited onto the substrate. Thus a pattern is formed on the substrate.
However, the above mentioned conventional methods of forming a fine pattern by employing a STM have to repeatedly carry out the step of moving a probe to align with an atom one by one and applying a voltage across the probe and the atom. Thus, the conventional methods have a problem that it takes too much time to form a pattern.
In addition, the generation of a tunneling current is greatly dependent on the shape of the probe, configuration and defects of a substrate, and contamination on a substrate etc. For instance, if there happens to exist contamination at a place where a probe is positioned, the probe may be destroyed or conditions for making a pattern are significantly varied.
For the above mentioned reasons, the conventional methods would take much time for forming a lot of patterns, and might be quite unstable in forming patterns. Furthermore, a probe of a STM has to be repositioned each time when the same pattern is to be formed in other areas, resulting in a problem that it is impossible to shorten a period of time for forming a lot of patterns.